Your observations are correct and just to expand a bit on them: a developer will build a price point that sells (obviously, the lower price points are selling better now and will probably be selling better in the future).
Building (as are most businesses) is all about sales velocity, internal rate of return (IRR) and minimizing risk. Pardee may make more on a larger house (in absolute dollars) but if it’s taking too long to sell, it erodes their IRR and increases their risk (their investment capital is out longer).
So, when a market starts to stall, builders come up with ways to cut costs and increase sales velocity, thereby increasing their IRRs back to the levels they need to make their financials looks good.
It’s all a balancing act and because of the long lead times, the fastest way to increase sales is decrease prices (build at a lower price point in this case or simply cut prices, usually with incentives so the actual sticker price doesn’t show a cut, if partway through a slow buildout) or if the market is roaring, build a bigger house with more profit.
Pardee shrunk the size of their homes because it would increase sales velocity, decrease the time to deliver those homes to the market, increase IRR (but not necessarily absolute dollar profit) and reduce their risk. All prudent business adjustments in a rapidly changing market.